What *that* Goldman Sachs report (https://lnkd.in/g6TW4mJV) gets right about #genAI It's become impossible to not hold a view on this report, at least in my professional circles. So here goes. First off, it's not really a single report, but a series of interviews and analyses from different perspectives, some of which contradict each other. I love that - too much certainty in this space underwhelms me. Read across, these various perspectives get three things right 1. That GenAI today costs too much (money, compute, energy) vs alternatives (the human brain or simpler automation) 2. That its main impact seems to be in making white-collar work more efficient rather than creating new products/ services (let alone solving humanity's big hairy problems) 3. That the technology is not yet reliable enough for even that efficiency-benefit to really kick in, at scale There is a huge amount of work underway on all three dimensions, though I am personally closest to the third bit. I remain bullish on AI more broadly (including GenAI), over the longer term.
+1, On cost, few reflections : I. It’s a problem of value creation and tracking. This takes time, but it’s feasible. II. Finding people who can think through hardware, software, and use cases without too many boundaries (or power struggles) to rethink your architecture and license costs is a rare skillset. III. Half of banking is still on-premise, and cloud migration plus data quality is likely a prerequisite to scale (and thus generate value).
Well said Shameek. For added context, sharing links to 2 related articles : 05 Apr. 2023 : GS -- Generative AI could raise global GDP by 7% https://www.goldmansachs.com/intelligence/pages/generative-ai-could-raise-global-gdp-by-7-percent.html 10 Nov. 2023 : Reuters - Goldman Sachs developing dozen generative AI projects https://www.reuters.com/technology/reuters-next-goldman-sachs-developing-dozen-generative-ai-projects-exec-2023-11-09/
Thanks for sharing, Shameek Kundu! I agree that the Goldman Sachs report offers valuable insights into the current state of #GenAI. While the technology is still evolving, I'm optimistic about its long-term potential, especially as it becomes more cost-effective and reliable. Kash Rangan's analogy to early SUN systems resonates with me, as it highlights how new technologies often take time to reach their full potential. Like many experts in the field, including Andrew Ng, Fei-Fei Li, Yann LeCun, Demis Hassabis, I believe we're just scratching the surface of what #AI can achieve.
Great report. Realism is much appreciated during this period of the hype cycle.
What I find interesting about genAI is that people (employees) will use it. It's ground-up. CTOs should be encouraging little personal gardens (walled) and seeing what crops come up, rather than top-down IT programs. I don't think those work with GenAI.
Linkedin Top Voice II Ex-McKinsey II GenAI Product and Growth leader in Banking, FinTech || CMO and Head of Data science Foodpanda (Unicorn) II Ex-CBO and Product leader Tookitaki
1moShameek Kundu thanks for sharing your perspective I will buy this if the author has ever build something esp. any AI product.